RepostedMeta logo

Meta · OA → Onsite

Rejected

Meta AI Native Engineer OA and Onsite: 1 Experience

Software engineer · Full-timeMaster's degreeReported Nov 19, 2026
Shared byjanebeingjane

A 2026 Meta AI Native Engineer loop: a banking OA, two AI-enabled coding rounds, behavioral, a music chart design and an agentic ticket triage design.

At a glance

Coverage: One candidate changing jobs, with a Master's degree, referred by a friend for a full-time AI Native Engineer role listed under general software engineering. The report covers the OA and the full onsite loop. It gives July to September 2026 as the interview period and went up on November 19, 2026.

What came up

The candidate calls both AI-enabled coding tasks common ones that others have reported before, and points to the two system designs as the rounds that went wrong (report 01).

Topic Reports Last seen
OA: a banking system ending in time-based lookups 1 of 1 Q3 2026
AI-enabled coding: a maze task 1 of 1 Q3 2026
AI-enabled coding: a card game 1 of 1 Q3 2026
System design: a music player with charts shared among friends 1 of 1 Q3 2026
AI system design: agentic ticket triage 1 of 1 Q3 2026

Questions we found

Algorithms / Data Structures

Question Reports Last seen
Banking system OA; the last part is like a time-based key-value store (practice) 1 Q3 2026

AI Coding

Question Reports Last seen
The maze task (practice) 1 Q3 2026
Card game 1 Q3 2026

System Design

Question Reports Last seen
Online music player where users share charts with friends (practice) 1 Q3 2026
Agentic ticket triage system (practice) 1 Q3 2026

1 detailed candidate timeline, loops, and screens

01: AI Native Engineer, OA then a five-round loop, not passed

Report date: November 19, 2026 · Interview period: 2026 Q3 · Role: AI Native Engineer, changing jobs · Degree listed: Master's · Channel: Referral · Candidate's rating: Negative experience, hard · Result: Did not pass

OA

  1. After a friend's referral, a recruiter got in touch and sent an OA first.
  2. The OA was a banking system split into parts. The last part was similar to a time-based key-value store.
  3. The candidate did well and got an invitation to the full loop.
  4. The report gives only the banking theme and the last part. FastPrep's practice version is Bank Accounts with Balance History, which includes looking up an account's balance at a past time.

Onsite: two AI-enabled coding rounds

  1. The candidate calls both tasks common ones that others have reported.
  2. One was the maze kind of task. FastPrep's practice version of Meta's maze task is Repair and Extend a Maze Solver.
  3. The other was a card game. The report gives only that name, so it can't be matched to one FastPrep exercise.

Onsite: behavioral

  1. One behavioral round; the report gives no questions from it.

Onsite: classic system design

  1. Design an online music player where users can share charts with their friends. FastPrep's practice version is Design Friend-Shared Charts for a Music Player.

Onsite: AI system design

  1. Design an agentic ticket triage system. FastPrep's practice version is Design an Agentic Support Ticket Triage System.

Result

The candidate felt neither system design went well and expected to fail, and did not pass.


Common questions

What is on the Meta AI Native Engineer OA?

In report 01, a banking system split into parts whose last part was like a time-based key-value store. The candidate did well and moved on to the loop.

What rounds are in the Meta AI Native Engineer onsite?

Report 01 had five: two AI-enabled coding rounds, one behavioral round, one classic system design and one AI system design.

What do the Meta AI-enabled coding rounds ask?

The candidate in report 01 got a maze task and a card game, and calls both common, often-reported questions.

Keep practicing

Sources: Anywhere useful interview info can be found.

0 total reactions. Your reactions: none.
0 comments

Comments

0

No comments yet — be the first to share what you know.